Recombination among HIV-1 strains to form chimeric genomes, a result of infection from multiple sources, has been found with unexpectedly high frequency. However, it is unclear whether it must result from near simultaneous acquisition of two HIV-1 infections (what we will term "co-infection"), or if true superinfection can occur, i.e., acquisition of a second virus following the establishment of primary infection and antiretroviral immune responses. If superinfection occurs, it would suggest that the immune responses generated by natural infection or vaccine candidates may not be sufficient to block new infections. Understanding whether, how often and when superinfection occurs is therefore of enormous importance to AIDS vaccine development. In addition, the consequences of multiple source infection (what we will call "dual infection") on the natural history of HIV infection have not been documented. In the proposed study, we will evaluate specimens from four longitudinal cohorts of individuals to determine: A) whether true superinfection occurs; and, B) if so, how frequently it occurs and under what circumstances; and C) whether dual infection affects the natural history of disease. Each cohort presents unique opportunities for gaining insight into these questions. They include: 1) Men and women participating in the University of Washington Primary Infection Cohort (PIC). These patients are sampled intensively over the first year of infection, and we will focus on those with sexual partners that are identified and engaged in follow-up, and who did not initiate HAART during the period of study. Analysis of this cohort will allow us to determine whether there is a transient window of susceptibility to dual infection; 2) A Male Partners Cohort (MPC) with established HIV infection prior to initiating their relationship, and who admit to having repeated engaged in unprotected anal sex. Having identified the viral strains in each sexual partner from the PIC and MPC cohorts, we can use highly sensitive, variant-specific probe sequences to identify low levels of dual infection at both early (PIC) and late (MPC) times after acquisition of primary infection; 3) The Multicenter AIDS Cohort Study (MACS) who have been followed at semiannual visits since 1984. This cohort will be uniquely valuable in determining the frequency of dual infection occurring throughout the disease process; and 4) HIV-1 seroconverters from a West African female prostitute cohort (WAC) with repeated, largely unprotected exposure to multiple HIV-1 subtypes. Analysis of this cohort will permit further evaluation of the hypothesis that dual infections are acquired early infection, and will uniquely permit evaluation of the emergence of dual infection with the same vs. different HIV-1 subtypes in a highly exposed and untreated population. In our studies, the emergence of co-infecting strains and chimeras will be identifed by genetic screening assays, including strain-specific amplification, heteroduplex tracking assay and DNA sequencing. The information obtained will be combined with analysis of T cell phenotypes, clinical status, plasma viral load and viral evolution trends to establish the relationships between dual infection, superinfection, chimera formation, and disease progression.